Gender differences in adult death registration in Kenya: associated factors and decomposition of the gap–a cross-sectional study

Background Civil registration systems in many low- and middle-income countries capture only a fraction of vital events, with deaths recorded less completely than births. Globally, female deaths are more often unregistered, yet evidence from Kenya is lacking. We examined gender differences in adult death registration across three Kenyan Health and Demographic Surveillance System (HDSS) sites. Methods We conducted a cross-sectional household survey of 1363 households reporting a recent death (2022–2023) across three HDSS sites in Nairobi, Kisumu, and Kilifi. Firth penalised logistic regression identified factors associated with registration, and Fairlie decomposition quantified the gender gap. Reported reasons for (non) registration were analysed descriptively. Findings Among 1055 adult deaths, registration was 22.8% overall and higher in males than females (26.1% versus 18.8%; p=0.006). Among females, registration odds were higher in those aged 45–64 than 18–44 years (AOR 2.13, 95% CI [1.04–4.47], p=0.038). Among males, odds were lower for single than married (AOR 0.33, 95% CI [0.13–0.78], p=0.011); a wealth gradient was evident from middle to richest quintile than poorest, and both genders' odds were associated with economic activity and time since burial. Observable factors explained 74.5% of the 7.43 percentage point gender gap, with injury/accident contributing 13.1%. Inheritance needs and social benefits were common reasons for registration; low awareness, perceived low benefit, and cost were common barriers. Interpretation Low adult death registration completeness in Kenya is driven by socioeconomic and administrative circumstances not gender-differentiated treatment by the registration system. Strengthening certification-registration, reducing costs, and targeted awareness campaigns could improve completeness and reduce disparities. Funding Wellcome Trust Senior Fellowship (#224272).